Howard almost made the biggest blunder of his life when
he tossed the building blocks of a wonder drug into
a laboratory trash bag 14 years ago. He had just created
a new molecule at Pfizer's research building in Groton,
Conn.-a molecule that had never existed before and that,
he hoped, might eventually prove useful in treating
schizophrenia. After running a test on it, Howard poured
the contents of a test tube through a filter and into
a flask. Then he threw out the filter and examined the
liquid left behind. It didn't seem to have any of the
properties he had expected. Oh well, he thought, chemicals
can do weird things.
The next day, however, it occurred to him that perhaps
he was looking in the wrong place. "It's a good
thing I hadn't taken the trash out," he says, "because
my compound was stuck in that filter."
Over the course of the next decade, Pfizer would invest
hundreds of millions of dollars in what Howard yanked
from the trash. Today, his compound is called Geodon.
It went on the market almost two years ago and, by most
accounts, it's one of the best medications available
to people with schizophrenia. Pfizer won't say exactly
how much money it poured into Geodon, but one estimate
says a successful drug can require an investment of
$800 million before the Food and Drug Administration
will approve it. There are thousands of failures for
every success-and somehow Pfizer must pay for them as
In a good year, Pfizer might win approval for one or
two brand-new products. It then has only a short time
to earn back what it has invested. Geodon's patent,
for instance, expires in March 2007, at which point
generic copies will begin to flood the pharmacy shelves.
By then, Pfizer will have had just six years to pay
for the huge costs that went into making Geodon. The
company estimates that it spends $100 million on research
and development per week. Within the next five years,
it will lose the patents for four of its eight most
Everybody loves wonder drugs, but nobody seems to want
to pay for them. One industry-sponsored poll revealed
that 90 percent of Americans believe the price of prescription
drugs is a "major problem" and 60 percent
support outright price controls. Few people appreciate
that innovation doesn't come cheap, especially when
drugs must pass through expensive regulatory hurdles
meant to test their safety and effectiveness. Wonder
drugs may seem like miracles, but they don't rain down
upon us like manna. Producing even one requires a monumental
Perhaps nothing can illustrate this point better than
a story, and Geodon's story begins with the problem
of schizophrenia. "It's one of the worst mental
diseases out there," says Steven Romano, a Pfizer
psychiatrist. Schizophrenics suffer from hallucinations,
delusions, and social withdrawal, among other symptoms.
Some 2 million Americans have the disease, and many
of them can't hold down a regular job: John Nash, the
Princeton mathematician portrayed by Russell Crowe in
the movie A Beautiful Mind, may be this country's best-known
schizophrenic. "It has exacerbations and remissions,
but it's chronic-it's always there, and there's no cure,"
The earliest drugs for schizophrenia could control
the hallucinations and delusions, but only at the cost
of grisly side effects, such as severe muscular tremors.
During the 1970s and 1980s, scientists began looking
for alternatives, and Geodon had its origins in this
broad effort. By the early 1990s, new products that
didn't make muscles go haywire were on the market, but
these had their own downsides, such as sharp weight
gain. As a result, many schizophrenics were getting
their main symptoms under control, only to quit their
meds because they didn't like the side effects.
Harry Howard was one of several Pfizer employees in
the 1980s charged with producing compounds that might
improve the treatment of schizophrenics. "Discovery"
scientists like Howard spend their days assembling molecules
that have never existed before and then seeing how they
behave in test tubes. "This part of the process
is all about creativity," says Stevin Zorn, who
led much of the Geodon research. "There's no cookbook
on how to make a successful compound. It's art as much
It's also drudgery. Even after Howard had constructed
the original Geodon molecule-code-named CP88059-his
creation faced long odds. "For every 150,000 compounds
our scientists create, we figure one of them will make
it to the marketplace," says Derek Leishman, a
pharmacologist with Pfizer in England. It's so rare
for a compound to make it to the shelves, in fact, that
many discovery scientists will spend their entire careers
working on compounds that never become approved drugs.
When a compound looks promising in the test tubes,
Pfizer scientists begin putting it into animals-mostly
mice, rats, and dogs. They want to make sure their proto-drug
isn't toxic. They're also looking for clues about how
it might act in humans. With CP88059, scientists were
hoping to spot muscle rigidity, a negative side effect
that had been associated with other successful antipsychotic
medications. But they found nothing of the sort, which
suggested that CP88059 might be a different breed of
drug. By this point, in 1989, it was time for Geodon
to "get canned."
This is an important milestone, as a compound moves
from one side of the R&D ampersand to the other.
It's called "getting canned" because the company
puts out a "Candidate Alert Notice" (CAN)-basically
an all- points bulletin announcing that the drug has
cleared its first set of obstacles. Yet the most challenging
phase still lies ahead: Fewer than 10 percent of canned
compounds ever make it to the marketplace.
The discovery scientists had produced a fine, white
powder that looked similar to flour. But one of the
great predicaments companies like Pfizer face consists
in moving from the pure science of compound creation
to the applied science of developing a form of the drug
that patients can use, such as a pill or capsule. The
final product must be able to survive on the shelf for
a couple of years, remain stable in hot and cold temperatures,
and deliver enough punch in a single dose that patients
won't have to pop pills a dozen times a day.
This step is harder than it sounds, but it's essential
to the pharmaceutical business-and it's in this area
that Pfizer made its first great contribution to humanity.
In 1928, Alexander Fleming found that a common airborne
mold known as penicillium could vanquish pneumonia and
scarlet fever-but only in a petri dish. He had enormous
trouble preserving the mold's active substance. It wasn't
until the Second World War that anybody was able to
produce penicillin in a useful form. Pfizer was the
company that figured out how to do it.
Pfizer faced this same challenge with Geodon-converting
a known substance into a product people could actually
use. "It took us two years to learn how to get
it into patients, and then a few more years of optimizing,"
says Bijan Rasadi, one of Pfizer's chief pharmaceutical
scientists. The entire development process took nine
A canned compound eventually graduates to clinical
testing, when it's put into humans for the first time.
The first phase of Geodon's clinical trials began in
1991 and, as first phases always do, it involved a few
dozen healthy subjects-non-schizophrenics who were willing
to serve as human guinea pigs. The point of these initial
tests isn't to see whether a drug works, but to determine
whether it's safe for human consumption. "Geodon's
first phase was uneventful," says Rachel Swift,
Pfizer's executive director of clinical development.
That was precisely the news she wanted to hear.
Once a drug's safety has been established in Phase
One, scientists next examine its efficacy. These Phase
Two trials typically involve several hundred people
affected by the problem the drug is supposed to address.
Researchers are generally confident that their drug
will have some effect, but they need to determine ideal
Geodon's Phase Two trials were also uneventful-perhaps
a bit too much so. Pfizer's R&D scientists saw that
Geodon was helping patients, but it didn't outperform
existing medications. If Geodon were ever to make it
to the marketplace, it would require yet more investment,
and at the end of the process there still had to be
a reasonable hope that it would generate revenue. A
redundant drug wouldn't. There was a strong case for
putting the brakes on Geodon. This is a fairly common
occurrence even for drugs that enter the second phase
of clinical testing.
Yet Pfizer was also hearing from doctors involved in
the trials that compliance was up. Geodon patients weren't
putting on the kind of weight associated with other
treatments. If this drug were able to delete that side
effect, the researchers thought, then perhaps they really